The present invention relates to semiconductor devices, and in particular, to semiconductor devices having an organic low dielectric constant material and processes for the manufacture thereof.
In an effort to increase the performance and speed of semiconductor devices, semiconductor device manufacturers have sought to reduce the linewidth and spacing of interconnects while minimizing the transmission losses and reducing the capacitative coupling of the interconnects. One way to diminish power consumption and reduce capacitance is by decreasing the dielectric constant (also referred to as xe2x80x9ckxe2x80x9d) of the insulating material, or dielectric, that separates the interconnects. Insulator materials having low dielectric constants are especially desirable, because they typically allow faster signal propagation, reduce capacitance and cross talk between conductor lines, and lower voltages required to drive integrated circuits.
Since air has a dielectric constant of 1.0, a major goal is to reduce the dielectric constant of insulator materials down to a theoretical limit of 1.0, and several methods are known in the art for reducing the dielectric constant of insulating materials. These techniques include adding elements such as fluorine to the composition to reduce the dielectric constant of the bulk material. Other methods to reduce k include use of alternative dielectric material matrices. Another approach is to introduce pores into the matrix.
Therefore, as interconnect linewidths decrease, concomitant decreases in the dielectric constant of the insulating material are required to achieve the improved performance and speed desired of future semiconductor devices. For example, devices having interconnect linewidths of 0.13 or 0.10 micron and below seek an insulating material having a dielectric constant (k) less than 3.
Currently silicon dioxide (SiO2) and modified versions of SiO2, such as fluorinated silicon dioxide or fluorinated silicon glass (hereinafter FSG) are used. These oxides, which have a dielectric constant ranging from about 3.5-4.0, are commonly used as the dielectric in semiconductor devices. While SiO2 and FSG have the mechanical and thermal stability needed to withstand the thermal cycling and processing steps of semiconductor device manufacturing, materials having a lower dielectric constant are desired in the industry.
Methods used to deposit dielectric materials may be divided into two categories: spin-on deposition (hereinafter SOD) and chemical vapor deposition (hereinafter CVD). Several efforts to develop lower dielectric constant materials include altering the chemical composition (organic, inorganic, blend of organic/inorganic) or changing the dielectric matrix (porous, non-porous). Table I summarizes the development of several materials having dielectric constants ranging from 2.0 to 3.5. (PE=plasma enhanced; HDP=high-density plasma) However, the dielectric materials and matrices disclosed in the publications shown in Table 1 fail to exhibit many of the combined physical and chemical properties desirable and even necessary for effective dielectric materials, such as higher mechanical stability, high thermal stability, high glass transition temperature, high modulus or hardness, while at the same time still being able to be solvated, spun, or deposited on to a substrate, wafer, or other surface. Therefore, it may be useful to investigate other compounds and materials that may be used as dielectric materials and layers, even though these compounds or materials may not be currently contemplated as dielectric materials in their present form.
Unfortunately, numerous organic SOD systems under development with a dielectric constant between 2.0 and 3.5 suffer from certain drawbacks in terms of mechanical and thermal properties as described above; therefore a need exists in the industry to develop improved processing and performance for dielectric films in this dielectric constant range. In addition, industry demands materials having demonstrated low dielectric constant extendibility, i.e. capable of being reduced to an even lower dielectric constant, e.g., from 2.7 to 2.5 to 2.2 to 2.0 and below.
Reichert and Mathias describe compounds and monomers that comprise adamantane molecules, which are in the class of cage-based molecules and are taught to be useful as diamond substitutes. (Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1993, Vol. 34 (1), pp. 495-6; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1992, Vol. 33 (2), pp. 144-5; Chem. Mater., 1993, Vol. 5 (1), pp. 4-5; Macromolecules, 1994, Vol. 27 (24), pp. 7030-7034; Macromolecules, 1994, Vol. 27 (24), pp. 7015-7023; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1995, Vol. 36 (1), pp. 741-742; 205th ACS National Meeting, Conference Program, 1993, pp. 312; Macromolecules, 1994, Vol. 27 (24), pp. 7024-9; Macromolecules, 1992, Vol. 25 (9), pp. 2294-306; Macromolecules, 1991, Vol. 24 (18), pp. 5232-3; Veronica R. Reichert, PhD Dissertation, 1994, Vol. 55-06B; ACS Symp. Ser.: Step-Growth Polymers for High-Performance Materials, 1996, Vol. 624, pp. 197-207; Macromolecules, 2000, Vol. 33 (10), pp. 3855-3859; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1999, Vol. 40 (2), pp. 620-621; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1999, Vol. 40 (2), pp. 577-78; Macromolecules, 1997, Vol. 30 (19), pp. 5970-5975; J. Polym. Sci, Part A: Polymer Chemistry, 1997, Vol. 35 (9), pp. 1743-1751; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1996, Vol. 37 (2), pp. 243-244; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1996, Vol. 37 (1), pp. 551-552; J. Polym. Sci., Part A: Polymer Chemistry, 1996, Vol. 34 (3), pp. 397-402; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1995, Vol. 36 (2), pp. 140-141; Polym, Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1992, Vol. 33 (2), pp. 146-147; J. Appl. Polym. Sci., 1998, Vol. 68 (3), pp. 475-482). The adamantane-based compounds and monomers described by Reichert and Mathias are preferably used to form polymers with adamantane molecules at the core of a thermoset. The compounds disclosed by Reichert and Mathias in their studies, however, comprise only one isomer of the adamantane-based compound by design choice. Structure A shows this symmetrical para-isomer 1,3,5,7-tetrakis[4xe2x80x2-(phenylethynyl)phenyl]adamantane: 
In other words, Reichert and Mathias in their individual and joint work contemplated a useful polymer comprising only one isomer form of the target adamantane-based monomer. A significant problem exists, however, when forming and processing polymers from the single isomer form (symmetrical xe2x80x9call-paraxe2x80x9d isomer) 1,3,5,7-tetrakis[4xe2x80x2-(phenylethynyl)phenyl]adamantane of the adamantane-based monomer. According to the Reichert dissertation (supra) and Macromolecules, vol. 27, (pp. 7015-7034) (supra), the symmetrical all-para isomer 1,3,5,7-tetrakis[4xe2x80x2-(phenylethynyl)phenyl]adamantane xe2x80x9cwas found to be soluble enough in chloroform that a 1H NMR spectrum could be obtained. However, acquisition times were found to be impractical for obtaining a solution 13C NMR spectrum.xe2x80x9d indicating that the all para isomer has low solubility. Thus, the Reichert symmetrical xe2x80x9call-paraxe2x80x9d isomer 1,3,5,7-tetrakis[4xe2x80x2-(phenylethynyl)phenyl]adamantane is insoluble in standard organic solvents and therefore, would not be useful in any application requiring solubility or solvent-based processing, such as flow coating, spin coating, or dip coating.
In our commonly assigned pending patent application PCT/US01/22204 filed Oct. 17, 2001 (claiming the benefit of our commonly assigned pending patent applications U.S. Ser. No. 09/545058 filed Apr. 7, 2000; U.S. Ser. No. 09/618945 filed Jul. 19, 2000; U.S. Ser. No. 09/897936 filed Jul. 5, 2001; and U.S. Ser. No. 09/902924 filed Jul. 10, 2001; and International Publication WO 01/78110 published Oct. 18, 2001), we discovered a composition comprising an isomeric thermosetting monomer or dimer mixture, wherein the mixture comprises at least one monomer or dimer having the structure correspondingly 
wherein Z is selected from cage compound and silicon atom; Rxe2x80x21, Rxe2x80x22, Rxe2x80x23, Rxe2x80x24, is Rxe2x80x25, and Rxe2x80x26 are independently selected from aryl, branched aryl, and arylene ether; at least one of the aryl, the branched aryl, and the arylene ether has an ethynyl group; and Rxe2x80x27 is aryl or substituted aryl. We also disclose methods for formation of these thermosetting mixtures. This novel isomeric thermosetting monomer or dimer mixture is useful as a dielectric material in microelectronics applications and soluble in many solvents such as cyclohexanone. These desirable properties make this isomeric thermosetting monomer or dimer mixture ideal for film formation at thicknesses of about 0.1 xcexcm to about 1.0 xcexcm.
We filed a patent application Serial No. 10/160,773 on even date herewith that claims a porous version of the preceding isomeric mixture.
Our International Patent Publication WO 01/78110 published Oct. 18, 2001 teaches in its background section that methods for introducing nanosized voids include physical blending or chemical grafting of thermostable or thermolabile portions. This publication""s invention is that nanosized voids may be introduced into dielectric materials by using cage structures such as adamantane or diamantane to achieve low dielectric constant material and defines low dielectric constant materials as having a dielectric constant of less than 3.0. However, this publication does not report any dielectric constant for its examples.
International Patent Publication WO 00/31183 teaches in its background section that although known porous thermoplastic materials had acceptable dielectric constants, the pores tended to collapse during subsequent high temperature processing and thus, the art teaches away from adding porosity to the cage structure that introduced nanosized voids in International Patent Publication WO 01/78110 published Oct. 18, 2001. In addition, U.S. Pat. Nos. 5,776,990; 5,895,263; 6,107,357; and 6,342,454 and U.S. Publication 2001/0040294 teach that although dielectric constants of 2.3-2.4 had been achieved at porosity levels less than about 20%, the pore content could not be further increased without comprising the small domain sizes and/or the non-interconnectivity of the pore structure. Similarly, U.S. Pat. Nos. 6,271,273; 6,156,812; 6,171,687; and 6,172,128 teach that the amount of the thermally labile monomer unit is limited to amounts less than about 30% by volume because if more than about 30% by volume of the thermally labile monomer is used, the resulting dielectric material has cylindrical or lamellar domains, instead of pores or voids, which lead to interconnected or collapsed structures upon removal, i.e., heating to degrade the thermally labile monomer units.
Although various methods are known in the art to lower the dielectric constant of a material, these methods have disadvantages. Thus, there is still a need in the semiconductor industry to a) provide improved compositions and methods to lower the dielectric constant of dielectric layers; b) provide dielectric materials with improved properties, such as thermal stability, glass transition temperature (Tg), modulus, and hardness; c) produce thermosetting compounds and dielectric materials that are capable of being solvated and spun-on to a wafer or layered material; and d) provide materials with demonstrated extendibility.
The present invention advantageously provides demonstrated extendibility so that semiconductor device manufacturers can use the present compositions for numerous generations of microchips. Also, the present invention provides for bonding of a porogen to a thermosetting component and thus, porogen movement is minimized and the possibility of pore aggregation is reduced.
In response to the need in the art and proceeding contrary to the wisdom in the art, we developed a composition comprising:
(a) thermosetting component comprising: (1) optionally monomer of Formula I 
xe2x80x83and (2) at least one oligomer or polymer of Formula II 
xe2x80x83where E is a cage compound; each of Q is the same or different and selected from aryl, branched aryl, and substituted aryl wherein said substituents include hydrogen, halogen, alkyl, aryl, substituted aryl, heteroaryl, aryl ether, alkenyl, alkynyl, alkoxyl, hydroxyalkyl, hydroxyaryl, hydroxyalkenyl, hydroxyalkynyl, hydroxyl, or carboxyl; G is aryl or substituted aryl where substituents include halogen and alkyl; h is from 0 to 10; i is from 0 to 10; j is from 0 to 10; and w is 0 or 1;
(b) porogen that bonds to the thermosetting component (a).
We also discovered a method of lowering the dielectric constant of a composition comprising: (a) thermosetting component comprising: (1) optionally monomer of Formula I above and (2) at least one oligomer or polymer of Formula II above where E, Q, G, h, I, and j are defined as above; and
(b) adhesion promoter comprising compound having at least bifunctionality wherein the bifunctionality may be the same or different and the first functionality is capable of interacting with the thermosetting component (a) and the second functionality is capable of interacting with a substrate when the composition is applied to the substrate
comprising the steps of:
bonding porogen to the thermosetting component;
decomposing the bonded porogen; and
volatilizing the decomposed porogen whereby pores form in the composition.